Morris Chang spent twenty-five years at Texas Instruments before the Taiwan government approached him in 1985 with a proposal: come to Taiwan and help build a domestic semiconductor industry. Chang, then 54 years old, accepted. He had a specific idea that no one had tried at commercial scale: a semiconductor company that only manufactured chips — no design, no product sales, no competition with its own customers. The 1987 founding capital of $220 million came from three sources: the Taiwan government through ITRI held 48%, Philips Electronics contributed 27.5%, and private Taiwanese investors provided the remaining 24.5%. The government stake was not a subsidy — it was a strategic investment in domestic industrial capability, and it provided TSMC with the credibility to attract customers who might otherwise have worried about a new, unproven foundry. The pure-play foundry concept took years to validate. The first major customers were small design firms that could not afford their own fabs. The business grew slowly through the early 1990s as the fabless model proved its commercial viability. TSMC listed on the New York Stock Exchange in 1994, providing access to international capital and establishing the company as a legitimate peer to US semiconductor companies. The 28-nanometer process node, launched in 2010, was the inflection point that established TSMC's technological leadership beyond dispute. NVIDIA, Qualcomm, and Apple moved manufacturing to TSMC's 28nm process and discovered that TSMC's yields, consistency, and technical support were superior to internal fab operations at every customer that had previously manufactured its own chips. From that point, the fabless model became the industry default for new chip designers, and TSMC's capacity became the scarce resource that determined how quickly the industry could grow.